Over the past few years, open radio access network (O-RAN) has gained wider acceptance from telecom service providers across the globe, as it has proved to be vital for efficient next-generation mobile network infrastructure. That said, service providers are reimagining networks. Several global telcos and vendors have initiated the process of transforming their O-RAN space by leveraging new-age technologies such as artificial intelligence (AI) and machine learning (ML), big data, edge computing and virtualisation.

These technologies are enabling operator networks to become more flexible and agile. Further, with advanced technologies, O-RAN not only improves network performance, but also enhances the efficiency and quality of service (QoS), and reduces network management and deployment costs.

A look at advanced technologies leveraged by network operators in the O-RAN space…

Automation through AI and ML

AI and ML essentially add intelligence to RAN, and thus, are widely adopted by telcos and vendors worldwide. Further, with the advent of 5G, there has been an increase in data complexity and traffic and to address this, industry players are shifting their focus from proprietary to virtualised or cloud-native network functions. Besides, the complexity of various 5G technologies, coupled with the shift towards O-RAN, can present several operational challenges for telecom operators. This, in turn, can disrupt the management of end-to-end performance. In this regard, AI can be critical in the management process of this complexity. Further, with the help of AI, telcos can deliver the quality of experience (QoE) demand by consumers and enterprises from latency-sensitive services and mobile broadband applications.

To this end, several companies are looking at adopting advanced technologies in the O-RAN space. For instance, in early 2021, Nokia and China Mobile completed live trials of AI for RAN applications on the carrier’s 4G and 5G network. The two companies tested an AI-based real-time user equipment (UE) traffic bandwidth forecast in Shanghai and automated network anomaly detection in Taiyuan. In addition, a RAN Intelligence Controller was deployed in edge cloud infrastructure. Further, in Shanghai, the trial confirmed that AI-based real-time UE traffic prediction accuracy exceeded 90 per cent in a live 5G network test.

Moreover, the use of AI technologies also involves disruptive changes to many existing software engineering, workflow stages and algorithm design. For instance, the development and operations (DevOps) flow requires extension with new data and ML model life-cycle management procedures. Further, network design, dynamic infrastructure optimisation and RAN algorithm or user optimisation requires re-engineering in a staged approach. To this end, O-RAN solutions can leverage AI and ML technologies in order to significantly improve operation automation through algorithms. This helps in significant enhancement of the operational efficiency. Going forward, various RAN features, which were traditionally manually programmed, will now rely on AI and automation capabilities to handle the increasing complexity of current and future networks.

Virtualisation through SDN and NFV

According to the O-RAN Alliance, RAN cloudification is one of the fundamental principles of the O-RAN architecture. Over the past few years, RAN virtualisation (vRAN) has emerged as a significant part of the RAN architecture roadmap. This offers various advantages to the mobile network operators (MNOs). Moreover, O-RAN is a major paradigm shift in RAN architecture that leverages software defined networking (SDN) and network function virtualisation (NFV) techniques by disaggregating the functions of a traditional RAN. O-RAN also implements SDN and NFV techniques in software and deploys them on independent cloud infrastructure, further connecting them using standardised interfaces. With this architecture, operators can accelerate service innovation and introduce intelligence in RAN control. This not only enables interoperability among RAN components from different sources, but also improves the supply chain security. In addition, introducing virtualisation in O-RAN also reduces network capex and opex costs.

That said, telecom operators and vendors are moving towards RAN virtualisation, which, in tandem, is benefiting vendors on cloud RAN (C-RAN) and vRAN architectures. For instance, in 2020, Ericsson announced C-RAN, a new offering to enable communications service providers to add greater flexibility and versatility to their networks. C-RAN by Ericsson is a cloud-native software solution, handling compute functionality in RAN. It will complement high-performing purpose-built baseband offerings in the Ericsson Radio System portfolio, giving service providers an optimal choice for any deployment scenario and need. Ericsson Cloud RAN will be released in stages, matching the service providers’ journey to complement their purpose-built 5G networks.

Further, the market segment deploying C-RAN architectures is among the first to virtualise RAN. Early adopters of vRAN also include in-building and in-venue segments and other stakeholders with available fibre fronthaul. For instance, these networks can be found in South Korea with KT, in Japan with NTT DOCOMO and SoftBank and in China with China Mobile, SK Telecom and LGU+. Besides, several operators in Europe are also running their vRAN trials, which include Deutsche Telekom, Orange France and Telefónica. Further, in the Middle East, Etisalat Mobily and Ooredoo have commercial vRAN networks.

Edge computing improving efficiency

As operators and vendors are focusing on virtualising their O-RAN deployments through SDN and NFV, edge computing becomes vital. While using a vRAN architecture, edge computing proves to be a more accessible and reasonable option, as RAN functions can be split between different aspects by network operators. For instance, vRAN is commonly split between the distributed unit (DU) and central unit (CU). While the CU may be on-site at the edge with a DU, it can also be centrally located at an aggregation site and interact with multiple DUs. Further, moving resources to the RAN edge can significantly reduce latency, as instead of processing data in the cloud, it can be done closer to the user. In addition, edge computing in O-RAN can also alleviate network congestion. Besides, it can help in the growth of autonomous vehicles, improve several experiences such as mobile gaming, and support dense internet of things deployments, among other use cases.

Strengthening analytics with big data

Big data analytics is helping network operators to enable network intelligence to O-RAN network elements across RAN, O-RAN controller, edge core and security gateway product suites. With big data, a large amount of complex, unstructured and structured data can be stored and processed at one place. This creates many opportunities for operators and helps them improve QoS for end-users and businesses. For instance, Parallel Wireless network intelligence software module integrates big data analytics with network optimisation via real-time, self-organising network, resulting in the improvement of end-user QoE. The company collects big data from all components across the network and other sources to form dynamic data lakes.

In sum 

Net, net, the trend towards the adoption of new-age technologies such as AI, automation, ML, cloud solutions, SDN, NFV and edge computing solutions are creating new market dynamics and opportunities for O-RAN. Platforms such as Facebook’s Telecom Infra Project will further fuel the innovation drive in the communications network architecture.